Subject: Re: VAXStation 3100 and MicroVAX 2000
To: Curt Sampson <curt@portal.ca>
From: Ken Wellsch <kcwellsc@math.uwaterloo.ca>
List: port-vax
Date: 10/20/1995 15:21:44
| So what's this bit of code doing in uvaxII_conf() in uvaxII.c?
| 
|         switch (cpu_type) {
| 	...
|         case VAX_410:
|                 strcpy(cpu_model,"MicroVAX 2000");
|                 break;

You'll note the ifdef's for the "uVAX-II" chipset.  the KA410 uses the
same CPU/FPU as the KA630 - which is likely why the are lumped together.

| Right. So what's involved in getting `everything around' running?
| How much coding would be necessary? Is adequate documentation easily
| available? Is this a project I should take up for my spare time,
| given that I know little about VAXen? :-)

Do you know how different the underlying implementation is?

There is a 16Kb memory buffer (going from memory as my detailed books are
in storage 1600 miles away) that is shared by the NCR 5380 SCSI chip (DEC
used it for the TK50 tape unit) and the Hitachi HD/FD controller (I forget
the number, something like 8115...).

Data transfers are made via these chips doing DMA to this buffer then the
CPU copying out/in to main memory.  The Lance? based EtherNet is the only
DMA peripheral as I recall.

If we had MOP going that would help a lot I think, since the farm of
vs2000's I've piled up are all without removable media.  All RD53 based.
I'm not sure how common this is for others, maybe they have a neato
tape unit, or the RX33 and an RD3x drive pair.  Either would give
access to removable media for debugging development.

Of course I am also presuming you'd talk through the serial port using
the standard 8 pin console cable from a uVAX-II as otherwise you have
to also deal with the video display which I've forgotten but would guess
is unique yet again.

A nice neat little package but enough unique key parts to make a serious
porting job I would guess necessary.